Search NASA⌕ Search

SEARCH · Search NASA

Results for “parallel, scaling”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 217 records · Page 12
All sourcesNASA NTRSDOE OSTI
NASA NTRS1991

Optimal processor assignment for pipeline computations

The availability of large scale multitasked parallel architectures introduces the following processor assignment problem for pipelined computations. Given a set of tasks and their precedence constraints, along with their experimentally determined individual responses times for different processor sizes, find an assignment of processor to tasks. Two objectives are of interest: minimal response given a throughput requirement, and maximal throughput given a response time requirement. These assignment problems differ considerably from the classical mapping problem in which several tasks share a processor; instead, it is assumed that a large number of processors are to be assigned to a relatively small number of tasks. Efficient assignment algorithms were developed for different classes of task structures. For a p processor system and a series parallel precedence graph with n constituent tasks, an O(np2) algorithm is provided that finds the optimal assignment for the response time optimization problem; it was found that the assignment optimizing the constrained throughput in O(np2log p) time. Special cases of linear, independent, and tree graphs are also considered.

Nicol, David M.↗
NASA NTRS1991

Response of composite materials to low velocity impact

Orthotropic and quasi-isotropic laminates clamped at the edges were impacted by an instrumented falling weight. Five matrix materials, comprising a baseline epoxy, two toughened thermosets, and amorphous and crystalline thermoplastics were studied. For each material, the projected damage area and residual compression strengths and strains were determined as a function of impact energy. Principal conclusions are: (1) incipient damage associated with a prominent load drop during the impact test seems to be decisive for residual properties; (2) subtle rate effects are present with some materials; and (3) results from the small scale tests parallel those in standard tests and may be adequate for screening purposes.

Srinivasan, K.↗
NASA NTRS1995

Real-time design with peer tasks

We introduce a real-time design methodology for large scale, distributed, parallel architecture, real-time systems (LDPARTS), as an alternative to those methods using rate or dead-line monotonic analysis. In our method the fundamental units of prioritization, work items, are domain specific objects with timing requirements (deadlines) found in user's specification. A work item consists of a collection of tasks of equal priority. Current scheduling theories are applied with artifact deadlines introduced by the designer whereas our method schedules work items to meet user's specification deadlines (sometimes called end-to-end deadlines). Our method supports these scheduling properties. Work item scheduling is based on domain specific importance instead of task level urgency and still meets as many user specification deadlines as can be met by scheduling tasks with respect to urgency. Second, the minimum (closest) on-line deadline that can be guaranteed for a work item of highest importance, scheduled at run time, is approximately the inverse of the throughput, measured in work items per second. Third, throughput is not degraded during overload and instead of resorting to task shedding during overload, the designer can specify which work items to shed. We prove these properties in a mathematical model.

Goforth, Andre↗
NASA NTRS2015

Laser Amplifier Development for the Remote Sensing of CO2 from Space

Accurate global measurements of tropospheric CO2 mixing ratios are needed to study CO2 emissions and CO2 exchange with the land and oceans. NASA Goddard Space Flight Center (GSFC) is developing a pulsed lidar approach for an integrated path differential absorption (IPDA) lidar to allow global measurements of atmospheric CO2 column densities from space. Our group has developed, and successfully flown, an airborne pulsed lidar instrument that uses two tunable pulsed laser transmitters allowing simultaneous measurement of a single CO2 absorption line in the 1570 nm band, absorption of an O2 line pair in the oxygen A-band (765 nm), range, and atmospheric backscatter profiles in the same path. Both lasers are pulsed at 10 kHz, and the two absorption line regions are sampled at typically a 300 Hz rate. A space-based version of this lidar must have a much larger lidar power-area product due to the approximately x40 longer range and faster along track velocity compared to airborne instrument. Initial link budget analysis indicated that for a 400 km orbit, a 1.5 m diameter telescope and a 10 second integration time, a approximately 2 mJ laser energy is required to attain the precision needed for each measurement. To meet this energy requirement, we have pursued parallel power scaling efforts to enable space-based lidar measurement of CO2 concentrations. These included a multiple aperture approach consists of multi-element large mode area fiber amplifiers and a single-aperture approach consists of a multi-pass Er:Yb:Phosphate glass based planar waveguide amplifier (PWA). In this paper we will present our laser amplifier design approaches and preliminary results.

Solid-State Laser Amplifier↗
NASA NTRS2016

High Energy, Narrow Linewidth 1572nm Eryb-Fiber Based MOPA for a Multi-Aperture CO2 Trace-Gas Laser Space Transmitter

Accurate global measurements of tropospheric CO2 mixing ratios are needed to study CO2 emissions and CO2 exchange with the land and oceans. NASA Goddard Space Flight Center (GSFC) is developing a pulsed lidar approach for an integrated path differential absorption (IPDA) lidar to allow global measurements of atmospheric CO2 column densities from space. Our group has developed, and successfully flown, an airborne pulsed lidar instrument that uses two tunable pulsed laser transmitters allowing simultaneous measurement of a single CO2 absorption line in the 1570 nm band, absorption of an O2 line pair in the oxygen A-band (765 nm), range, and atmospheric backscatter profiles in the same path. Both lasers are pulsed at 10 kHz, and the two absorption line regions are sampled at typically a 300 Hz rate. A space-based version of this lidar must have a much larger lidar power-area product due to the x40 longer range and faster along track velocity compared to airborne instrument. Initial link budget analysis indicated that for a 400 km orbit, a 1.5 m diameter telescope and a 10 second integration time, a 2 mJ laser energy is required to attain the precision needed for each measurement. To meet this energy requirement, we have pursued parallel power scaling efforts to enable space-based lidar measurement of CO2 concentrations. These included a multiple aperture approach consists of multi-element large mode area fiber amplifiers and a single-aperture approach consists of a multi-pass Er:Yb:Phosphate glass based planar waveguide amplifier (PWA). In this paper we will present our laser amplifier design approaches and preliminary results.

micropulse lidar↗
NASA NTRS2019

Molecular Dynamics Simulations of Silicon Carbide, Boron Nitride and Silicon for Ceramic Matrix Composite Applications

A comprehensive computational molecular dynamics study is presented for crystalline α-SiC (6H, 4H, and 2H SiC), β-SiC (3C SiC), layered boron nitride, amorphous boron nitride and silicon, the constituent materials for high-temperature SiC/SiC compositions. Large-scale Atomic/Molecular Parallel Simulator software package was used. The Tersoff Potential force field was utilized to evaluate their mechanical characteristics of most of the materials, and the Reax force field was used to model silicon when the Tersoff Potential did not provide accurate results. Their mechanical behaviors were evaluated at a strain rate of 10(exp 7)/s and the results agree with the experimental data in the literature. The results are foundational for linking constituent behavior to composite performance, particularly when test data is unavailable or suspect.

Aluko, Olanrewaju↗
DOE OSTI2021

Improving scalability of electronic structure code for molecular simulations in the presence of environment

A scalable density functional electronic code with Gaussian basis set, called UTEP-NRLMOL, is developed to perform simulations of molecular systems in the presence of the environment with particular attention to the memory requirements. In the electronic structure calculations, the memory and computation time are proportional to the number of atoms. Memory requirements for density functional calculations scale as N*N, where N is the number of atoms. While the recent advances in HPC offer platforms with large numbers of cores, the limited amount of memory available on a given node and poor scalability of the electronic structure codes hinder their efficient usage of these platforms. We have introduced new scaling and parallelization paradigms using MPI-3 shared-memory functionality combined with usage of sparse algebra and storage of matrices in sparse format. This extends the range of applicability of the UTEP-NRLMOL code to large systems over 10,000 atoms, or using up to 67,000 basis functions, and making use of HPC architectures using over 6,000 processors utilizing all available cores. We have also interfaced code with effective fragment potential and polarizable continuum model libraries. The code was used in simulations of several applications which are published in reputed scientific journals.

74 ATOMIC AND MOLECULAR PHYSICS↗
NASA NTRS1993

Symposium on Parallel Computational Methods for Large-scale Structural Analysis and Design, 2nd, Norfolk, VA, US

Computing speed is leaping forward by several orders of magnitude each decade. Engineers and scientists gathered at a NASA Langley symposium to discuss these exciting trends as they apply to parallel computational methods for large-scale structural analysis and design. Among the topics discussed were: large-scale static analysis; dynamic, transient, and thermal analysis; domain decomposition (substructuring); and nonlinear and numerical methods.

Storaasli, Olaf O.↗
DOE OSTI2017

A parallel orbital-updating based plane-wave basis method for electronic structure calculations

Highlights: • Propose three parallel orbital-updating based plane-wave basis methods for electronic structure calculations. • These new methods can avoid the generating of large scale eigenvalue problems and then reduce the computational cost. • These new methods allow for two-level parallelization which is particularly interesting for large scale parallelization. • Numerical experiments show that these new methods are reliable and efficient for large scale calculations on modern supercomputers. - Abstract: Motivated by the recently proposed parallel orbital-updating approach in real space method , we propose a parallel orbital-updating based plane-wave basis method for electronic structure calculations, for solving the corresponding eigenvalue problems. In addition, we propose two new modified parallel orbital-updating methods. Compared to the traditional plane-wave methods, our methods allow for two-level parallelization, which is particularly interesting for large scale parallelization. Numerical experiments show that these new methods are more reliable and efficient for large scale calculations on modern supercomputers.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
NASA NTRS1989

Parallel-vector solution of large-scale structural analysis problems on supercomputers

A direct linear equation solution method based on the Choleski factorization procedure is presented which exploits both parallel and vector features of supercomputers. The new equation solver is described, and its performance is evaluated by solving structural analysis problems on three high-performance computers. The method has been implemented using Force, a generic parallel FORTRAN language.

Storaasli, Olaf O.↗
DOE OSTI2018

Understanding scale-dependent soft-error behavior of scientific applications

Analyzing application fault behavior on large-scale systems is time-consuming and resource-demanding. Currently, researchers need to perform fault injection campaigns at full scale to understand the effects of soft errors on applications and whether these faults result in silent data corruption. Both time and resource requirements greatly limit the scope of the resilience studies that can be currently performed. In this work, we propose a methodology to model application fault behavior at large scale based on a reduced set of experiments performed at small scale. We employ machine learning techniques to accurately model application fault behavior using a set of experiments that can be executed in parallel at small scale. Our methodology drastically reduces the set and the scale of the fault injection experiments to be performed and provides a validated methodology to study application fault behavior at large scale. We show that our methodology can accurately model application fault behavior at large scale by using only small scale experiments. In some cases, we can model the fault behavior of a parallel application running on 4,096 cores with about 90% accuracy based on experiments on a single core.

Kestor, Gokcen G.↗
NASA NTRS2018

The Role of the Parallel Electric Field in Electron-Scale Dissipation at Reconnecting Currents in the Magnetosheath

We report observations from the Magnetospheric Multiscale satellites of reconnecting current sheets in the magnetosheath over a range of outofplane "guide" magnetic field strengths. The currents exhibit nonideal energy conversion in the electron frame of reference, and the events are within the ion diffusion region within close proximity (a few electron skin depths) to the electron diffusion region. The study focuses on energy conversion on the electron scale only. At low guide field (antiparallel reconnection), electric fields and currents perpendicular to the magnetic field dominate the energy conversion. Additionally, electron distributions exhibit significant nongyrotropy. As the guide field increases, the electric field parallel to the background magnetic field becomes increasingly strong, and the electron nongyrotropy becomes less apparent. We find that even with a guide field less than half the reconnecting field, the parallel electric field and currents dominate the dissipation. This suggests that parallel electric fields are more important to energy conversion in reconnection than previously thought and that at high guide field, the physics governing magnetic reconnection are significantly different from antiparallel reconnection.

Wilder, F. D.↗
DOE OSTI2020

The Case for Strong Scaling in Deep Learning: Training Large 3D CNNs with Hybrid Parallelism

Here, we present scalable hybrid-parallel algorithms for training large-scale 3D convolutional neural networks. Deep learning-based emerging scientific workflows often require model training with large, high-dimensional samples, which can make training much more costly and even infeasible due to excessive memory usage. We solve these challenges by extensively applying hybrid parallelism throughout the end-to-end training pipeline, including both computations and I/O. Our hybrid-parallel algorithm extends the standard data parallelism with spatial parallelism, which partitions a single sample in the spatial domain, realizing strong scaling beyond the mini-batch dimension with a larger aggregated memory capacity. We evaluate our proposed training algorithms with two challenging 3D CNNs, CosmoFlow and 3D U-Net. Our comprehensive performance studies show that good weak and strong scaling can be achieved for both networks using up to 2K GPUs. More importantly, we enable training of CosmoFlow with much larger samples than previously possible, realizing an order-of-magnitude improvement in prediction accuracy.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2006

CAISSON: Interconnect Network Simulator

Cray response to HPCS initiative. Model future petaflop computer interconnect. Parallel discrete event simulation techniques for large scale network simulation. Built on WarpIV engine. Run on laptop and Altix 3000. Can be sized up to 1000 simulated nodes per host node. Good parallel scaling characteristics. Flexible: multiple injectors, arbitration strategies, queue iterators, network topologies.

network simulation↗
DOE OSTI2022

Low-synch Gram–Schmidt with delayed reorthogonalization for Krylov solvers

The parallel strong-scaling of iterative methods is often determined by the number of global reductions at each iteration. Low-synch Gram-Schmidt algorithms are applied here to the Arnoldi algorithm to reduce the number of global reductions and therefore to improve the parallel strong-scaling of iterative solvers for nonsymmetric matrices such as the GMRES and the Krylov-Schur iterative methods. In the Arnoldi context, the factorization is "left-looking" and processes one column at a time. Among the methods for generating an orthogonal basis for the Arnoldi algorithm, the classical Gram-Schmidt algorithm, with reorthogonalization (CGS2) requires three global reductions per iteration. A new variant of CGS2 that requires only one reduction per iteration is presented and applied to the Arnoldi algorithm. Delayed CGS2 (DCGS2) employs the minimum number of global reductions per iteration (one) for a one-column at-a-time algorithm. The main idea behind the new algorithm is to group global reductions by rearranging the order of operations. DCGS2 must be carefully integrated into an Arnoldi expansion or a GMRES solver. Numerical stability experiments assess robustness for Krylov-Schur eigenvalue computations. Performance experiments on the ORNL Summit supercomputer then establish the superiority of DCGS2 over CGS2.

97 MATHEMATICS AND COMPUTING↗